Exploring the composition of lithic assemblages in Mesolithic south-eastern Norway
Isak Roalkvam1,✉
19 April, 2021
1 University of Oslo
✉ Correspondence: Isak Roalkvam <isak.roalkvam@iakh.uio.no>
1 Introduction
The analyses conducted here employs multivariate exploratory statistics using the entirety of the lithic assemblages associated with a larger number of Mesolithic sites located in south-eastern Norway. This is done to identify latent patterns and structure in the relationship between these, with the ultimate aim of identifying behaviourally induced variation in their composition across time and space. However, the composition of the assemblages can be expected to be determined by a multitude of factors (e.g. Dibble et al., 2017), ranging from the impact of natural formation processes, to various and intermixed behavioural aspects such as purpose, duration, frequency, and group sizes at visits to the sites. Furthermore, the assemblages are also likely to be impacted by variation in lithic technology, artefact function, use-life and discard patterns, as well as access to raw materials. Finally, analytical and methodological dimensions relating to survey, excavation and classification practices are also fundamental to how the assemblages are defined. Consequently, the analysis conducted here is done from an exploratory perspective, where all of these factors should be seen as potential contributors to any observed pattern.
Although each individual assemblage can potentially have been impacted by an infinitude of effects that might skew an archaeological interpretation, this does not preclude the applicability of inductive analyses aimed at revealing overarching structure in the data, without imposing overly complex analytical frameworks that attempt to account for these particularities (Bevan, 2015). Structure that can be revealed from considering all of the assemblages in aggregate can constitute a step in an iterative analytical chain that ultimately aims to tease apart such aspects from the multitude of factors that have shaped the composition of the assemblages, and should be of value to subsequent in-depth studies of any individual site. The most immediate danger of the approach outlined here is rather to be overly naive in the causal significance and cultural importance that is ascribed to any identified pattern. As such, the main aim of this analysis is to compare the results with findings reported in previous literature concerned with the Mesolithic in southern Norway that have been based on more informally driven methods, and have the generation of new hypotheses as a possible outcome. To this end, the analysis follows two analytical avenues.
1.1 Archaeological context and material
The 54 coastal sites chosen for analysis here have a relatively limited geographical distribution in south-eastern Norway (fig:dist)A]. The sites were excavated as part of four larger excavation projects that all took place within the last 15 years (solheim2013b?; jaksland2014a?; melvold2014b?; reitan2014l?; solheim2017b?). The sites included in the analysis consists of all Mesolithic sites excavated in conjunction with the projects with assemblages holding more than 100 artefacts. The institution responsible for all of these excavations was the Museum of Cultural History in Oslo. This has led to a considerable overlap in the archaeological personnel involved, and comparable excavation practices across the excavations. Furthermore, with these projects, major efforts were made to standardise how lithic artefacts were to be classified at the Museum of Cultural History. As a result, this should reduce the amount of artificial patterning in the data incurred by discrepancies in the employed systems for categorisation (Clark and Riel-Salvatore, 2006; Dibble et al., 2017). In this setting, for example, bias could potentially follow from the fact that two of the projects have sites with relatively contemporaneous dates (Jaksland, 2014, see also (fig:dist)B; solheim2013b?). Any project-dependent classification practice could as a consequence lead to an exaggeration of chronological differences between the assemblages. While this is difficult to fully account for, I do believe that the relative contemporaneity of the excavation projects, as well as the overlap in excavation and classification practices should minimise the above-mentioned effects, making the data-set a good candidate for exploratory data analysis.
Figure 1.1: A) Spatial and B) temporal distribution of the sites chosen for analysis. The date ranges follow those given in the original excavation reports
A defining characteristic of the Norwegian Mesolithic is that a clear majority of the known sites are located in coastal areas. Furthermore, these coastal sites, including those treated here, appear to predominantly have been located on or close to the contemporary shoreline when they were in use (Breivik et al., 2018; Solheim et al., 2020). In south-eastern Norway, this pattern is combined with a continuous regression of the shoreline, following from isostatic rebound (Romundset et al., 2018; e.g. Sørensen, 1979). The fairly rapid shoreline displacement means that the sites tend not to have retained their strategic or ecologically beneficial shore-bound location for long periods of time (cf. perreault2019?). Consequently, the shore-bound settlement, combined with the rapid shoreline displacement has resulted in a relatively high degree of spatial separation of cumulative palimpsests, to follow the terminology of Bailey (2007), while shoreline displacement combined with shore-bound settlement allows for a relatively good chronological control of these accumulation events. In other parts of the world, a higher degree of spatial distribution means that while the physical separation of material can help delineate discrete events, this typically comes at the cost of loosing temporal resolution, as any stratigraphic relationship between the events is lost (Bailey, 2007). However, as the rate of isostatic rebound has varied throughout the Mesolithic in the region, and local topography and bathymetry will have impacted how rapidly a site lost its shore-bound location, this effect is not evenly distributed in time and space. In the earliest part of the Mesolithic, the displacement rate within the study area would have been around as much as 8.8 cm/year, falling to around 0.5 cm/year in the Late Mesolithic (Sørensen et al., 2014). Thus, while relative sea-level change appears to have reduced the degree of mixing that has occurred in the assemblages, this could vary depending on when and where they were in use, in turn potentially reducing the degree to which their composition can be directly compared - a point that is returned to in the discussion of the results.
The data analysed here is based on the classification of the 54 site assemblages done for the original excavation reports, and consists of 48 variables representing different debitage and tool types. While the classification practices for the excavation projects were standardised to an extent, there are some instances where time was allocated to identify additional artefact sub-categories aimed at answering specific research questions. Some categories in the original reports have therefore been combined in the dataset. This for example pertains to the category narrow-blades, which is defined as blades of width between 8 and 12 mm. In the reports, this was only separated from (macro-)blades (width \(\ge\) 12 mm) and micro-blades (width \(\le\) 8 mm) for some of the sites. This was combined with the blade category here. Furthermore, the artefact data has been divided into flint and non-flint materials. Flint does not occur indigenously in Norway, and is only available as nodules that have been transported and deposited by retreating and drifting ice (e.g. berg-hansen1999?). This means that the distribution and quality of the flint has been impacted by a diverse set of factors relating to climatic and geographical factors such as, but not limited to, topographic variability, shoreline morphology and ocean currents (eigeland2015?). Consequently, although flint is treated as a unified category here, there is reason to believe that this variability has had consequences for raw-material accessibility and technological choices (eigeland2015?). Furthermore, a general sentiment is that an increase in the use of non-flint materials signifies an increased adaption to local environments, and possibly a reduction in mobility. However, the various non-flint raw materials that have been lumped together here can have quite disparate properties, where fine-grained cryptocrystalline materials are often used as a substitute or supplement to flint, while other, coarser materials are usually associated with the production of axes and other macro tools. Given this differentiated use, the retained debitage and tool categories are expected to still reflect the differentiated use associated with different raw-materials. An important benefit of combining all non-flint materials is that this reduces the dependency on whether or not these have been correctly and consistently categorised (cf. frivoll2017?).
In a series of case studies from Northern Norway, Bølviken et al. (1982) were among a group of scholars often attributed the popularisaton of correspondence analysis (CA) in wider archaeologcal circles (Baxter, 1994). In southern Norway, however, the use of multivariate statistics in Mesolithic research is to my knowledge limited to (solheim2013a?) use of CA to analyse the assemblage data from eight sites excavated in connection with a development-led project. As these sites, which are included in this study, have relatively contemporaneous dates, the Solheims analysis was mainly aimed at identifying potential variation in the
Previous studies of lithic artefacts in Norway have typically had a focus on chronological trends associated with formal tool types (Reitan, 2016; e.g. helskog1976?), involved refit-studies of the entirety of individual assemblages (Skar and Coulson, 1986), analyses concerned with technological processes associated with certain sub-categories of the site inventories (Damlien, 2016; Solheim et al., 2020), or in-depth analyses of handful of sites. Some studies have attempted more comprehensive evaluations of entire lithic assemblages associated with larger groups of sites, mainly to get at potential site types and associated mobility patterns among Early Mesolithic sites (Breivik, 2020; breivik2016?; viken2018a?). However, only a single study has employed a multivariate quantitative framework in the treatment of the assemblages (cf. naeroy2018?). In sum, therefore, previous studies are typically either limited to a small number of sites, to a subset of the inventories, or to qualitatively and subjectively driven methods.
There have been a few studies that have conducted comprehensive analyses of artefact inventories and site features to get at variation in Mesolithic site types (Breivik, 2020; Viken, 2018). However, the narratively driven analyses of these data makes the weighting of different variables unclear and subjectively defined, and several of the studies cited here ultimately draw their conclusions based on an arguably over-constraining trichotomy of site types. As Binford originally clarified, a point that has also been retiterated by several authors since; these categories are best understood as extremes on a continuous scale.
1.2 Methods
The first part of the analysis involves employing the method of correspondence analysis (CA), using the lithic count data as classified for the original excavation reports. The purpose of this exercise is to evaluate the degree to which the composition of the assemblages align with patterns that have been suggested by earlier studies that have employed more informally driven methods. This assumes that the artefact categories employed in Norwegian Stone Age archaeology are, at least to a certain extent, behaviorally meaningful. However, the approach taken is partially informed by the so-called Frison Effect, which highlights the fact that the shape of lithics studied by archaeologists is that which they had when they where discarded - the artefacts can have had long and complex use-lives in which they took on a multitude of different shapes. Several scholars have built on this to argue that morphological variation in retouched lithics is not predominantly the result of the intention of the original knapper to reach some desired end-product, but rather that what is commonly categorised as discrete types of artefacts by archaeologists can instead in large part be related to variable degrees of modification through use and rejuvination (e.g. Barton, 1991; Dibble, 1995) . Consequently, several artefact categories have here been collapased for the CA. This for example pertains to tool types such as scrapers, burins, drills and knives and indeterminate artefacts with retouch. That these categories are internally consistent and categorically exclusive in terms of fulfilled purpose, is at best a dubious proposition, in turn rendering their contribution as discrete analytic units potentially entirely misleading. These have all been combined into the single category “small flint tools.” (A full overview of the aggregated variables and their constituent parts is provided in the supplementary material). While aggregating artefact groups in this manner could potentially subsume important variation, it does also reduce the possibility that any drawn conclusions are not simply the result of employing erroneous units of analysis. An underlying assumption is therefore that the retained categories represent artefact categories that have fulfilled different purposes or are related to different technological processes. While ultimately intuitive in nature, it does seem reasonable to assume that for example large non-flint stone tools such as axes, adzes, chisels, clubs and hatches, here categorised as non-flint macro tools, have fulfilled different purposes than the previously mentioned small flint tools.
However, the fact that for the most part we lack even a most basic understanding of what any individual lithic object has been used for does have major implications that the above-outlined analysis does not take into account, and renders it difficult to align any identified patterns with specific behavioural dimensions (Dibble et al., 2017). For example, a vast amount of artefacts defined as debitage are likely to have fulfilled the function of tools, and both debitage and formal tool types could have had various different purposes and had a multitude of shapes throughout their use-life. While use-wear analysis could potentially offer a way to identify what artefacts were used for towards the end of their use-life, these kinds of analyses are extremely time-consuming and are therefore typically only conducted on a smaller number of artefacts that have already been selected for analysis based on their shape (e.g. solheim2018?). Thus, while these analyses can potentially get at in-group variation pertaining to the end-state of a group of artefacts, they do not tell us whether or not their classification as a unified group is meaningful in the first place (Dibble et al., 2017). As a consequence, the second part of the analysis employs a suite of measures developed for the classification of lithic assemblages developed with these inferential limitations in mind. The logic behind these measures are founded on an understanding of technology as being organised along a continuum ranging between curated and expedient (Binford, 1979). An expedient technological organisation pertains to the situational production of tools to meet immediate needs, with little investment of time and resources in modification and rejuvination, resulting in high rates of replacement. Curated technological organisation, on the other hand, has been defined as related to manufacture and maintenance of tools in anticipation of future use, the transport of these artefacts between places of use, and the modification and rejuvenation of artefacts for different and changing situations (Binford, 1979). Although precisely how degrees of retouch is measured may vary, the empirical correspondent to a curated technological organisation is therefore typically defined by high degrees of retouch, as this is commonly seen as a means of extending the use-life or realising the potential utility of a tool by repeated rejuvination and modification of edges (e.g. Bamforth, 1986; Dibble, 1995; Shott and Sillitoe, 2005).
The continuum between curated and expedient technology has in turn been related to land-use and mobility strategies (Clark and Barton, 2017; e.g. Parry and Kelly, 1987), as well as raw-material quality and availability (e.g. Andrefsky, 1994; Smith, 2015). Higher degree of mobility would mean a higher dependency on the artefacts and the material people could bring with them, and dimensions such as weight, reliabilty, repairability, and the degree to which artefacts could be manipulated to fulfill a wide range of tasks are therefore assumed to have been factors of concern. From this it follows that the empirical expectation for short-term camps is a higher relative frequency of retouched artefacts, and a lower overall density of lithics (Clark and Barton, 2017). More time spent in a single location, on the other hand, is in this framework assumed to lead to better control of raw-material availability, and to allow for the accumulation of material resources, which should in turn lead to an expedient technological organisation and reduce the necessity for extensive retouch. The material expectation for lower degree of mobility is therefore relatively high density of lithics and a low relative frequency of retouched artefacts. Additional empirical expectations following from this is a higher number of cores and unretouched flakes and blades (Clark and Barton, 2017).
While the concept of curation and its empirical and behavioural implications have been widely discussed Shott (1996) , following not least from the ambiguous definition first put forward by Binford (1979), some concrete operationalisation of the terms have been suggested. Through a series of studies, have shown that the relationship between lithic volumetric density and relative frequency of retouched artefacts provide consistent and comparable results across a wide range of chronological and cultural context which are taken to reflect . This is the main measure employed here. Some additional variables suggested by Bicho and Cascalheira (2020) to be related to mobility patterns are subsequently also explored using principal components analysis (PCA), following the continuous operationalisations of these measures (cf. Baxter, 1994, p. 100).
A note should also be made on the fact that a couple of variables that are sometimes invoked for the classification of sites in terms of associated mobility pattern are omitted here. This pertains to number of site features such as fireplaces, cooking pits and dwelling structures which has been omitted as taphonomic loss is likely to lead to a chronological bias in their preservation. Similarly, the number of activity areas, effectively number artefact clusters, however defined, has also been disregarded. This follows most notably from the fact that the impact of post-depositional processes at Stone Age sites in Norway is arguably understudied. This pertains for example to the impact of bioturbation in the form of three-throws, which can can have a detrimental effect on the original distribution of artefacts, and which can be expected to have been relatively frequent on several of the sites treated here (Jørgensen, 2017; darmark2018e?).
Within this framework, amount of retouch has been seen as a empirical indicator of degree of curation. While various methods for measuring the curation associated with individual tools have been forwarded, the approach taken here focuses on the assemblage considered as a whole.
Here retouch i At logistical basecamps, the extended time spent in a single location would imply a better control of raw material availability, and allow for longer trips to retrieve these. Higher availability of lithic raw materials should in turn reduce the necessity for extensive retouch, meaning the material expectation for logistical basecamps would be relatively high density of lithics and a low relative frequency of retouched artefacts. From this it follows that the empirical expectation for short-term camps is a higher relative frequency of retouched artefacts, and a lower overall density of lithics (Clark and Barton, 2017).
2 Theory/calculation
Higher frequency of secondarily worked lithics
3 Results
Figure 3.1: Correspondence analysis using artefact count data. A) Object map, B) Variable map.
Figure 3.1 displays the CA using the lithic count data. While no clear-cut clusters can be readily delineated, the general impression from the plots is that a chronological dimension is associated with the patterning in the data. This is indicated by the general transition across the colour scale in the row plot (Figure 3.1A), combined with the fact that the two first dimensions of the CA accounts for as much as 80.53 % of the inertia or variance in the data. The earliest sites tend to be located in the upper right corner of plot A, with increasingly younger sites towards the bottom along the second dimension. Although fewer in number, the sites from the later parts of the Mesolithic are drawn out along the first dimension of the plot, and are not as impacted by the second dimension as the more numerous older sites.
The column plot (Figure 3.1B) reveals that the earliest sites are characterised by the flint artefact categories microburins, projectiles, as well as macro tools and associated debitage. It is also interesting that these sites to larger extent are characerised by core fragments, both in flint and non-flint materials, as opposed to the cores themselves. The non-flint material on the earliest sites appears to be centered around the production of projectiles, as both the projectiles themselves and non-flint blades are important constituents of the assemblages at these sites.
The first dimension, which is pulling some of the later sites towards the right of the plot, is mainly defined by macro tools and associated debitage in non-flint materials that are negatively correlated with more flint dominated assemblages and non-flint projectiles. While the outer end of the first dimensions is dominated by later Mesolithic sites associated with axe production in non-flint materials, the later sites occur along the entire dimension, indicating that while these axe production sites are a feature of the later Mesolithic, there is marked variation among these later sites. Although the sample size is quite strained and the discussion of finer chronological points might not be warranted, the first dimension does appear to be of of less importance for the absolute latest sites, as indicated by their location to the left of the plot. This could indicate that specialised axe production sites disappear towards the end of the Mesolithic, a notion that would be in line with previous research (e.g. eigeland2015?).
In sum, the results of the CA does appear to align reasonably well with previous research that has been based on more informal analyses of artefact types, indicating that the employed artefact categories are capturing some behaviorally meaningful patterning that changes over time. However, as several authors have pointed out in relation to the application of CA for seriation purposes, ‘time is not the only dimension.’ This is evident in the present data as well, highlighted by the influence of the later Mesolithic axe production sites, which it would be reasonable to assume corresponds to a more pervasive cultural change than a purely typo-chronological development of artefact shape. Unpicking and aligning the signficance of these chronological patterns with any specific behavioural dimensions using the CA results is, however, another task entirely. This follows most clearly from the fact that for the most part we lack even a most basic understanding of what any individual lithic object has been used for, leaving the significance of the employed units of analysis unclear (Dibble et al., 2017).
In addition, Nedre Hobekk 2 represents a somewhat curious case in that its assemblage is dominated by axe production in metarhyolite. The use of metarhyolite is typically seen as a feature of the end of the EM and the MM, but is evidently not as prominent a part of other sites that are contemporaneous with Nedre Hobekk 2.
4
Reducing the number of such inferential leaps by aggregating artefact groups could potentially subsume important variation, but it does also reduce the possibility that the conclusions of any analysis is not simply the result of the employed units of analysis, and an overly naive trust in the fact that these units reflect any behaviourally meaningful categories. This realisation has led to a wide range of outside Scandinavian Mesolithic research, meaning a range of empirical measures associated with various mobility patterns are available.
It has, however, been suggested that access to raw-materials can be a more important decider for assemblage composition than mobility patterns , and that variable workability of these materials will impact factors such as the relative frequency of tools to debitage in the assemblages (e.g. Andrefsky, 1994; Manninen and Knutsson, 2014). In addition to attempts at accounting for this analytically (see below), the limited geographical distribution of the sites should alleviate the issue somewhat, as the overall availability of raw materials could be expected to be more comparable through time than if a wider region was under consideration.
Of these the most well-estbalished is the. base don the assumption that, this measure and has been shown to display a meaningful measure across a wide range of temporal, geographical and cultural contexts.
Figure 4.1 displays the WABI as found for the site data. This indicates a negative correlation between the two variables (r = -0.5) and a general tendency for younger sites to be associated with a higher volumetric density of lithics and a lower relative frequency of secondarily worked lithics than older sites. Following Clark and Barton (2017), this would be in line with a general transition from a expedient to curated lithic technology. Variable raw-material availability can also impact these measures as (Manninen and Knutsson, 2014). Variation in raw-material use over time could consequently mean that these patterns could be an effect of the available raw material rather than mobility strategies. However, as is evident in when considering both plots in 4.1, the pattern is evident for both the lithics considered in aggregate and for the flint data specifically. Although the distribution is slightly more spread out along the y-axis than the x-axis in the flint data, indicating that differences in volumtric density of flint is not as clearly chronologically, the general pattern is the same.
Figure 4.1: Relative frequency of secondarily worked lithics plotted against the volumetric density of artefacts (artefact count / excavated m3) for A) All lithics, B) Flint. The logarithm is taken to base 10 on all axes.
Figure 4.2: PCA.
Figure 4.2 displays a principle components analysis using variations of the continuous measures for degree of mobility as operationalised by Bicho and Cascalheira (2020). While the investigation performed by Bicho and Cascalheira (2020) indicates that the relative frequency of chips, cores, and blanks might be more sensitive to mobility patterns, they also found that the relationship between volumetric density of lithics and the frequency of retouched artefacts gave a reasonable estimation of mobility in their data. hese last dimensions are capturing the most variation in the dataset presented here. Thus, if frequency of secondarily worked artefacts is accepted as a proxy for degree of mobility, these findings would consequently be in line with previous research into the Mesolithic of Norway, indicating that earlier sites are associated with higher degree of mobility than sites from later phases.
| Component | Eigenvalue | Variance (%) | Cumulative variance (%) |
|---|---|---|---|
| 1 | 2.43 | 48.59 | 48.59 |
| 2 | 0.97 | 19.48 | 68.07 |
| 3 | 0.96 | 19.21 | 87.28 |
| 4 | 0.63 | 12.63 | 99.92 |
| 5 | 0.00 | 0.08 | 100.00 |
(#fig:var_contrib)Contribution of variables to components.
(#fig:var_cor)Plots showing the correlation between the variables
5 Discussion
The consistent nature of the negative correlation between LVD and relative retouch frequency across this study (r = -0.50), Clark and Barton (2017) (r = -0.50) and Bicho and Cascalheira (2020) (r = -0.48) is striking. While this result is perhaps not surprising given the previous and plentiful indications of the same trends, these results do speak to the cross-cultural and cross-temporal applicability of the measure, and consequently the potential it holds for larger scale comparative studies. The findings indicated by the WABI, as well as by the negative correlation between relative frequency of primarily and secondarily worked lithics, do align more clearly with previous suggestions concerning the overall mobility patterns in Mesolithic Norway. However, the question then becomes: What is the significance of the negatively correlated variables of relative frequency of cores and chips? Given that these are more or less orthogonal to the WABI variables on the first and second component, this would indicate that if one of these variable pairs is taken to reflect the of expedient and curated assemblages, it should follow that the other variable pairs is not correlated with these. As the entire temporal range of the sites is spread out between the extremes of the chip and core variables, one possibility is that these are indicative of site types that are temporally ubiquitous. In attempt to determine if this might be the case, it was decided to introduce locational data from a previous study that in addition to surveyed sites involved the analysis of the sites treated here Roalkvam (2020). This found that the most consistently important locational variable for the sites in the region was the exposure of the sites to the elements, and failed to identify any diachronic variation in this pattern. As it has previously been proposed that sites located in areas exposed to wind and waves and with large overview of surrounding areas is a characteristic of hunting stations or similar, one could envisage that variablity in site functionality indicated by the core to chip correlation might be reflected in variation in degree of exposure.
This is arguably more established than the variables included bysimilarities between the variable plots of Bicho and Cascalheira (2020) and the variables indicated in is striking. This pertains to the negative correlation between core and chip frequency, and the negative correlation between primary and secondarily worked lithics. This last
These measures are taken from the framework of whole assemblage behavioural index, as developed by , and is aimed to get at behavioural dimensions relating to mobility patterns among hunter-gatherers.
6 References
6.0.1 Colophon
This report was generated on 2021-04-19 12:25:39 using the following computational environment and dependencies:
#> ─ Session info ───────────────────────────────────────────────────────────────
#> setting value
#> version R version 4.0.5 (2021-03-31)
#> os Linux Mint 19.3
#> system x86_64, linux-gnu
#> ui X11
#> language en_US
#> collate en_US.UTF-8
#> ctype en_US.UTF-8
#> tz Europe/Oslo
#> date 2021-04-19
#>
#> ─ Packages ───────────────────────────────────────────────────────────────────
#> package * version date lib source
#> abind 1.4-5 2016-07-21 [1] CRAN (R 4.0.3)
#> assertthat 0.2.1 2019-03-21 [1] CRAN (R 4.0.3)
#> backports 1.2.0 2020-11-02 [1] CRAN (R 4.0.3)
#> bitops 1.0-6 2013-08-17 [1] CRAN (R 4.0.3)
#> bookdown 0.21 2020-10-13 [1] CRAN (R 4.0.3)
#> broom 0.7.3 2020-12-16 [1] CRAN (R 4.0.3)
#> callr 3.5.1 2020-10-13 [1] CRAN (R 4.0.3)
#> car 3.0-10 2020-09-29 [1] CRAN (R 4.0.3)
#> carData 3.0-4 2020-05-22 [1] CRAN (R 4.0.3)
#> cellranger 1.1.0 2016-07-27 [1] CRAN (R 4.0.3)
#> checkmate 2.0.0 2020-02-06 [1] CRAN (R 4.0.3)
#> class 7.3-18 2021-01-24 [4] CRAN (R 4.0.3)
#> classInt 0.4-3 2020-04-07 [1] CRAN (R 4.0.3)
#> cli 2.2.0 2020-11-20 [1] CRAN (R 4.0.3)
#> cluster 2.1.1 2021-02-14 [4] CRAN (R 4.0.3)
#> colorspace 1.4-1 2019-03-18 [1] CRAN (R 4.0.3)
#> crayon 1.3.4 2017-09-16 [1] CRAN (R 4.0.3)
#> curl 4.3 2019-12-02 [1] CRAN (R 4.0.3)
#> data.table 1.13.4 2020-12-08 [1] CRAN (R 4.0.3)
#> DBI 1.1.0 2019-12-15 [1] CRAN (R 4.0.3)
#> dbplyr 2.0.0 2020-11-03 [1] CRAN (R 4.0.3)
#> desc 1.2.0 2018-05-01 [1] CRAN (R 4.0.3)
#> devtools 2.3.2 2020-09-18 [1] CRAN (R 4.0.3)
#> digest 0.6.27 2020-10-24 [1] CRAN (R 4.0.3)
#> dplyr * 1.0.2 2020-08-18 [1] CRAN (R 4.0.3)
#> DT 0.16 2020-10-13 [1] CRAN (R 4.0.3)
#> e1071 1.7-4 2020-10-14 [1] CRAN (R 4.0.3)
#> ellipsis 0.3.1 2020-05-15 [1] CRAN (R 4.0.3)
#> evaluate 0.14 2019-05-28 [1] CRAN (R 4.0.3)
#> factoextra * 1.0.7 2020-04-01 [1] CRAN (R 4.0.3)
#> FactoMineR * 2.4 2020-12-11 [1] CRAN (R 4.0.3)
#> fansi 0.4.1 2020-01-08 [1] CRAN (R 4.0.3)
#> farver 2.0.3 2020-01-16 [1] CRAN (R 4.0.3)
#> flashClust 1.01-2 2012-08-21 [1] CRAN (R 4.0.3)
#> forcats * 0.5.0 2020-03-01 [1] CRAN (R 4.0.3)
#> foreign 0.8-81 2020-12-22 [4] CRAN (R 4.0.3)
#> fs 1.5.0 2020-07-31 [1] CRAN (R 4.0.3)
#> generics 0.1.0 2020-10-31 [1] CRAN (R 4.0.3)
#> ggmap 3.0.0 2019-02-05 [1] CRAN (R 4.0.4)
#> ggplot2 * 3.3.2 2020-06-19 [1] CRAN (R 4.0.3)
#> ggpubr 0.4.0 2020-06-27 [1] CRAN (R 4.0.3)
#> ggrepel 0.9.1 2021-01-15 [1] CRAN (R 4.0.3)
#> ggsignif 0.6.0 2019-08-08 [1] CRAN (R 4.0.3)
#> ggsn 0.5.0 2019-02-18 [1] CRAN (R 4.0.4)
#> glue 1.4.2 2020-08-27 [1] CRAN (R 4.0.3)
#> gt * 0.2.2 2020-08-05 [1] CRAN (R 4.0.3)
#> gtable 0.3.0 2019-03-25 [1] CRAN (R 4.0.3)
#> haven 2.3.1 2020-06-01 [1] CRAN (R 4.0.3)
#> here 1.0.0 2020-11-15 [1] CRAN (R 4.0.3)
#> highr 0.8 2019-03-20 [1] CRAN (R 4.0.3)
#> hms 0.5.3 2020-01-08 [1] CRAN (R 4.0.3)
#> htmltools 0.5.0 2020-06-16 [1] CRAN (R 4.0.3)
#> htmlwidgets 1.5.2 2020-10-03 [1] CRAN (R 4.0.3)
#> httr 1.4.2 2020-07-20 [1] CRAN (R 4.0.3)
#> jpeg 0.1-8.1 2019-10-24 [1] CRAN (R 4.0.3)
#> jsonlite 1.7.1 2020-09-07 [1] CRAN (R 4.0.3)
#> KernSmooth 2.23-18 2020-10-29 [1] CRAN (R 4.0.3)
#> knitr 1.30 2020-09-22 [1] CRAN (R 4.0.3)
#> labeling 0.4.2 2020-10-20 [1] CRAN (R 4.0.3)
#> lattice 0.20-41 2020-04-02 [1] CRAN (R 4.0.3)
#> leaps 3.1 2020-01-16 [1] CRAN (R 4.0.3)
#> lifecycle 0.2.0 2020-03-06 [1] CRAN (R 4.0.3)
#> lubridate 1.7.9.2 2020-11-13 [1] CRAN (R 4.0.3)
#> magrittr 2.0.1 2020-11-17 [1] CRAN (R 4.0.3)
#> maptools 1.0-2 2020-08-24 [1] CRAN (R 4.0.3)
#> MASS 7.3-53.1 2021-02-12 [4] CRAN (R 4.0.3)
#> Matrix 1.3-2 2021-01-06 [4] CRAN (R 4.0.3)
#> memoise 1.1.0 2017-04-21 [1] CRAN (R 4.0.3)
#> mgcv 1.8-33 2020-08-27 [1] CRAN (R 4.0.3)
#> modelr 0.1.8 2020-05-19 [1] CRAN (R 4.0.3)
#> munsell 0.5.0 2018-06-12 [1] CRAN (R 4.0.3)
#> nlme 3.1-152 2021-02-04 [4] CRAN (R 4.0.3)
#> openxlsx 4.2.3 2020-10-27 [1] CRAN (R 4.0.3)
#> patchwork * 1.1.0 2020-11-09 [1] CRAN (R 4.0.3)
#> pillar 1.4.7 2020-11-20 [1] CRAN (R 4.0.3)
#> pkgbuild 1.1.0 2020-07-13 [1] CRAN (R 4.0.3)
#> pkgconfig 2.0.3 2019-09-22 [1] CRAN (R 4.0.3)
#> pkgload 1.1.0 2020-05-29 [1] CRAN (R 4.0.3)
#> plyr 1.8.6 2020-03-03 [1] CRAN (R 4.0.3)
#> png 0.1-7 2013-12-03 [1] CRAN (R 4.0.3)
#> prettyunits 1.1.1 2020-01-24 [1] CRAN (R 4.0.3)
#> processx 3.4.4 2020-09-03 [1] CRAN (R 4.0.3)
#> ps 1.4.0 2020-10-07 [1] CRAN (R 4.0.3)
#> purrr * 0.3.4 2020-04-17 [1] CRAN (R 4.0.3)
#> R6 2.5.0 2020-10-28 [1] CRAN (R 4.0.3)
#> Rcpp 1.0.5 2020-07-06 [1] CRAN (R 4.0.3)
#> readr * 1.4.0 2020-10-05 [1] CRAN (R 4.0.3)
#> readxl 1.3.1 2019-03-13 [1] CRAN (R 4.0.3)
#> remotes 2.2.0 2020-07-21 [1] CRAN (R 4.0.3)
#> reprex 0.3.0 2019-05-16 [1] CRAN (R 4.0.3)
#> rgdal 1.5-18 2020-10-13 [1] CRAN (R 4.0.3)
#> RgoogleMaps 1.4.5.3 2020-02-12 [1] CRAN (R 4.0.4)
#> rio 0.5.26 2021-03-01 [1] CRAN (R 4.0.4)
#> rjson 0.2.20 2018-06-08 [1] CRAN (R 4.0.4)
#> rlang 0.4.10 2020-12-30 [1] CRAN (R 4.0.4)
#> rmarkdown 2.5 2020-10-21 [1] CRAN (R 4.0.3)
#> rnaturalearth * 0.1.0 2017-03-21 [1] CRAN (R 4.0.5)
#> rprojroot 2.0.2 2020-11-15 [1] CRAN (R 4.0.3)
#> rstatix 0.6.0 2020-06-18 [1] CRAN (R 4.0.3)
#> rstudioapi 0.13 2020-11-12 [1] CRAN (R 4.0.3)
#> rvest 0.3.6 2020-07-25 [1] CRAN (R 4.0.3)
#> sass 0.2.0 2020-03-18 [1] CRAN (R 4.0.3)
#> scales 1.1.1 2020-05-11 [1] CRAN (R 4.0.3)
#> scatterplot3d 0.3-41 2018-03-14 [1] CRAN (R 4.0.3)
#> sessioninfo 1.1.1 2018-11-05 [1] CRAN (R 4.0.3)
#> sf * 0.9-7 2021-01-06 [1] CRAN (R 4.0.4)
#> sp 1.4-5 2021-01-10 [1] CRAN (R 4.0.4)
#> stringi 1.5.3 2020-09-09 [1] CRAN (R 4.0.3)
#> stringr * 1.4.0 2019-02-10 [1] CRAN (R 4.0.3)
#> testthat 3.0.0 2020-10-31 [1] CRAN (R 4.0.3)
#> tibble * 3.0.4 2020-10-12 [1] CRAN (R 4.0.3)
#> tidyr * 1.1.3 2021-03-03 [1] CRAN (R 4.0.4)
#> tidyselect 1.1.0 2020-05-11 [1] CRAN (R 4.0.3)
#> tidyverse * 1.3.0 2019-11-21 [1] CRAN (R 4.0.3)
#> units 0.6-7 2020-06-13 [1] CRAN (R 4.0.3)
#> usethis 2.0.1 2021-02-10 [1] CRAN (R 4.0.4)
#> vctrs 0.3.7 2021-03-29 [1] CRAN (R 4.0.4)
#> withr 2.3.0 2020-09-22 [1] CRAN (R 4.0.3)
#> xfun 0.19 2020-10-30 [1] CRAN (R 4.0.3)
#> xml2 1.3.2 2020-04-23 [1] CRAN (R 4.0.3)
#> yaml 2.2.1 2020-02-01 [1] CRAN (R 4.0.3)
#> zip 2.1.1 2020-08-27 [1] CRAN (R 4.0.3)
#>
#> [1] /home/isak/R/x86_64-pc-linux-gnu-library/4.0
#> [2] /usr/local/lib/R/site-library
#> [3] /usr/lib/R/site-library
#> [4] /usr/lib/R/libraryThe current Git commit details are:
#> Local: master /home/isak/phd/dialpast_r/dialpastrepository
#> Remote: master @ origin (https://github.com/isakro/dialpastrepository.git)
#> Head: [175f4d4] 2021-04-17: Set up Zotero web with RStudio